As semiconductor device features are reduced in size, conventional processing techniques (e.g., photolithography) are unable to directly meet the size requirements. The concept of pitch can be used to describe the sizes of features of a semiconductor device. Pitch may be defined as a distance between an identical point in two adjacent features when the pattern includes repeating features, as in arrays. These features are conventionally separated by spaces between adjacent features, which spaces are subsequently filled by a material, such as an insulator. As a result, pitch can be viewed as the sum of the width of a feature and of the width of the space on one side of the feature separating that feature from the adjacent feature. However, due to factors such as optics and radiation wavelength, photolithography techniques have a minimum pitch below which a particular photolithographic technique cannot reliably form features. Thus, the minimum pitch of a photolithographic technique is an obstacle to continued feature size reduction.
“Pitch doubling” or “pitch multiplication” is a technique that may be used to form features smaller than is possible by conventional photolithography technology. While pitch is actually reduced by this technique, the reduction in pitch is conventionally referred to as “pitch doubling” or, more generally, “pitch multiplication.” Thus, conventionally, “multiplication” of pitch by a certain factor actually involves reducing the pitch by that factor. The conventional terminology is retained herein.
In one method of pitch doubling, a feature is formed by conventional photolithography and a spacer is formed on sidewalls of the feature. Material from the spacer is removed from horizontal surfaces (e.g., a top of the feature, a floor of a space between adjacent features), leaving the spacers only along the sidewalls of the feature. The feature is removed, leaving two spacers for every one feature originally formed by photolithography (e.g., one spacer on each of two opposing sidewalls of the feature). The spacers form a pattern, which is transferred into an underlying material. Material underlying the spacers is retained, while material underlying an area between the spacers is removed to faint features under each spacer in a desired pattern. Alternatively or additionally, material may be formed (e.g., deposited) between the spacers, between features underlying the spacers, or within gaps and trenches formed under the spacers. Thus, a number of features can essentially be doubled in a given area, compared to conventional photolithography techniques.
If a metal pattern is formed by disposing a metal material within gaps and trenches formed under the spacers, the metal material may be subjected to an abrasive removal process (e.g., chemical-mechanical polishing (CMP)) to remove excess metal material. Such a process may cause so-called “dishing” in larger metal features, such as metal pads, in which central portions of the larger metal features are thinner than peripheral portions thereof. Dishing can lead to reliability issues and even failures of devices incorporating structures resulting from such a process.
Pitch doubling techniques involve an undesirable number of process acts to arrive at a final pattern. Alternative, improved methods for fabricating features of dimensions below resolution limits of photolithography are desirable.